Heart Team Recommendations Conflict with Original CAD Treatment

In a cross-sectional study among CAD patients, investigators found that 30.3% of heart team treatment recommendations diverged from original consultation.

A new study finds significant disparities between heart team recommendations and those of original treating interventional cardiologists for patients with multivessel coronary artery disease (CAD).

The study, led by Michael Tsang, MD, MSc, and colleagues and funded by the Hamilton Health Sciences New Investigator Fund, consisted of a cross-sectional study to determine the extent of differences in multivessel CAD treatment decisions between heart teams and original treating interventional cardiologists. Their findings also indicated that in cases when both group opinions differed, the heart team tended to have a greater frequency of internal discordance.The investigators noted that it is currently unclear whether heart team reviews generally lead to improved clinical outcomes, yet the results of this study indicate that it may be necessary to identify which patients would benefit from such an approach.

At the outset of the study, the investigators selected 245 out of 771 screened patients who met specific criteria for multivessel CAD, which they defined as stenosis of 70% or more in 3 epicardial vessels, stenosis of at least 1.5 mm in their branches, or stenosis in 2 epicardial coronary vessels with involvement of the proximal left anterior descending artery.

The recruitment process included an interview and treatment recommendation from an original treating interventional cardiologist.

Another 60 patients from their database with similar distribution of original treatment recommendations were retrospectively added to the data for analysis.

The investigators constructed 8 three-member heart teams, each consisting of 1 interventional cardiologist, 1 cardiovascular surgeon, and 1 noninvasive cardiologist. Each member was blind to their other team members and their decisions, so as to mitigate any social influences on decision-making, and decisions were made asynchronously and independently. Members were able to view the other decisions only after their independent recommendations were submitted—a change in decisions was then allowed.

Of the 245 selected patients, the teams evaluated 237 cases through a virtual heart team interface. Each team received 6 randomized sets of 5 cases. The investigators determined the heart team recommendation to be either the unanimous or majority decision.

The 3 recommendations considered among the original treating interventional cardiologists and heart teams were coronary artery bypass grafting (CABG), percutaneous coronary intervention, and medical therapy.

Results showed that 71 differences (30.3%; 95% CI, 24.5-36.7) occurred between both types of recommenders.

Tsang and colleagues used the Cohen x statistic to determine the overall agreement between both treatment cases. A paired analysis between original treatment decision and heart team decisions showed an x coefficient of .478 (95% CI, 0.336 - 0.540, P = .006), which indicates moderate agreement.

Further analysis revealed that agreement between the heart team decision and original treatment agreement was often associated with a unanimous recommendation among the heart team (109 of 163 cases [66.9%]), compared with discordance between both recommendations (28 of 71 cases [39.4%], P < .001). 

The heart team interventional cardiologist often agreed with the original testing cardiologist when both decisions were concordant with each other (138 of 163 cases [84.7%]) versus when both original treatment and heart team treatment disagreed with each other (14 of 71 cases [19.7], P < .001).

Tsang and team have noted the implications of such conflicting treatment decisions, especially in uncertain multivessel CAD cases.

“If use of the heart team approach were found to be associated with improvements in outcomes, selection of these cases a priori would likely require a scoring tool that uses common clinical characteristics (eg, age, frailty, cognitive dysfunction, and SYNTAX score) to quantify the therapeutic dilemma,” they wrote. “Because multivessel CAD accounts for approximately 25% to 60% of patients with CAD, such a tool would have wide applicability.”

The study, “Heart Team vs Interventional Cardiologist Recommendations for Multivessel Coronary Artery Disease,” was published online in JAMA Network Open.